Prostate cancer is the most common cancer among men in the United States and men diagnosed with this disease are faced with a difficult decision regarding treatment. With the advent of PSA testing, tumors are being diagnosed that may have gone undetected in the past. There is a lack of uniform opinion among physicians on the most appropriate therapy a man should receive. Aggressive therapies (e.g. radical prostatectomy and radiation) may provide the highest likelihood of cure; but the cost may be an impairment of sexual, urinary, and/or bowel function. Gleason score and pre-diagnostic PSA levels have been used as predictors of tumor aggressiveness, but they are not sufficient, especially for younger men with an intermediate grade tumor, to predict the long term aggressiveness of their tumor. Therefore, more specific biomarkers are needed. One promising area of research concerns the level of androgen receptor (AR) expression. The androgen-signaling axis is important in the development of the prostate gland and is associated with risk of prostate cancer as well as with progression of disease. We plan to perform immunohistochemistry to measure AR expression in both tumor and peritumoral cells at the time of initial diagnosis and to extract germline DNA from normal tissue to assay CAG and GGC microsatellite repeats in the AR gene. These biomarker results will be linked with indicators of disease progression, therapy and clinical characteristics and it will be determined through multivariate analysis if they can improve the prediction of progression when added to currently existing measures. The archived tumor tissue to be studied has been previously obtained from 290 men who were diagnosed with localized disease in 1994-95 and have been followed for progression for at least 5 years. Information is available on their therapy, clinical characteristics, and recurrence/progression (from medical records and self-reported questionnaires). The sample includes 140 whites, 78 African-Americans, and 72 Hispanics. Based on self-reports, about 22% (65) have progressed. The sample size will permit differences of 20% in prevalence of a biomarker between the progressors and nonprogressors to be detected as well as an increased risk of progression as low as 2.5. This study has implications for assisting men in making the right treatment choice, improving long-term quality of life, and reducing unnecessary therapy. It will also lay the groundwork for larger studies examining. [unreadable] [unreadable] [unreadable]